Liao and Chen Bot Stud (2017) 58:15 DOI 10.1186/s40529-017-0171-0

ORIGINAL ARTICLE Open Access Investigation of floristic similarities between and terrestrial ecoregions in Asia using GBIF data Chi‑Cheng Liao1* and Chih‑Hui Chen2

Abstract Background: Floristic compositions of non-endemic plants of continental islands were related to the neighboring continents because non-endemic plant species had historically migrated to continental islands from source areas. This study attempts to identify source areas of a continental island by means of floristic analysis and to assess pos‑ sible migration routes on the basis of geographical distribution ranges of plants. Large quantities of angiosperm data records were downloaded from the Global Biodiversity Information Facility (GBIF). Similarity index and cluster analy‑ sis were used to identify the floristic similarities among 22 geographical localities of Taiwan (GLTs) and 34 terrestrial ecoregions in Asia. Geographical distribution ranges of non-endemic angiosperm species in Taiwan (NEASTs) were evaluated to mirror the possible migration routes from different source areas to Taiwan. Results: There are 3275 angiosperm species in Taiwan derived from the dataset of GBIF. Among them, 847 are endemic and 2428 are NEASTs. Geographical distribution ranges of the 2428 NEASTs were categorized into 7 distri‑ bution groups. They were widely distribution from equator to Siberia (27 species), tropical ecoregions (345 species), tropical and subtropical ecoregions (663 species), tropical to temperate ecoregions (591 species), subtropical ecore‑ gions (265 species), subtropical to temperate ecoregions (387 species), and temperate ecoregions (150 species). Results of similarity indices and cluster analysis demonstrated that high floristic similarities were observed among GLTs at lowland and southern Taiwan and tropical and subtropical ecoregions in Asia. GLTs at high mountains were assumed to have floristic similarity with temperate ecoregions in Asia, whereas the assumption was not supported by our analysis. It is partly because of that angiosperms with tropical and subtropical distributions extend their ranges from low to high elevations in Taiwan. Conclusions: Subtropical ecoregions at southern and tropical ecoregions at Indochina were more important than temperate ecoregions on playing source areas of NEASTs. Geographical distribution ranges of NEASTs implied that most of the NEASTs were probably migrated from topical or subtropical ecoregions of Asian continent to Taiwan. Keywords: Angiosperm, Asia, Continental island, Floristic similarity, Geographical distribution, Taiwan, Terrestrial ecoregion

Background et al. 2009; Kreft et al. 2008; Krupnick et al. 2009; Mal- Oceanic islands possess disproportionately high plant colm et al. 2006; Waldren et al. 1995). On the contrary, species richness and numbers of endemic taxa (Bramwell continental islands possess relatively lower proportion of and Caujapé-Castells 2011; Brooks et al. 2002; Carlquist endemic taxa and floras of continental islands are closely 1967; Chen and He 2009; Cowie and Holland 2006; Kier related to floras of neighboring continents (Bramwell and Caujapé-Castells 2011; Fernández-Palacios et al.

*Correspondence: [email protected] 2011; Heaney 1991; Hsu and Wolf 2009). In , 1 Department of Life Science, Chinese Culture University, 55, Hwa‑Kang plant species richness of continental islands are closely Road, Yang‑Ming‑Shan, Taipei 11114, Taiwan, ROC related to the Eurasian continents and neighboring Full list of author information is available at the end of the article

© The Author(s) 2017. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Liao and Chen Bot Stud (2017) 58:15 Page 2 of 17

regions (Chiang and Schaal 2006; Hiramatsu et al. 2001; to northern areas and from low to high elevations in Tai- Nakamura et al. 2009; Ota 1998; Setoguchi et al. 2008). wan (Chen 1957). Diverse climatic conditions in Taiwan The extent to which non-endemic species of continental are influencing the plant distributions and differentiation islands are related to the neighboring regions remains of floristic compositions within the island (Chiou et al. elusive. 2010; Su 1984, 1985). Six geographical areas was divided Taiwan locates at the eastern border of Eurasian con- in Taiwan in terms of climatic environments, plant dis- tinent and was formed by the collision between Luzon tributions and floristic compositions (Su 1984, 1985). Arc and Eurasian continent during 2–3 million years ago Floristic differentiation among six geographical areas in (Chen and Liu 1996; Hsieh et al. 2006; Liew and Hsieh Taiwan leads to two assumptions. The first is that dif- 2000; Teng 1990, 2007; Voris 2000; Wei 2002; Zeng 1993). ferent floristic composition of geographical areas in Tai- Land bridge connections between Taiwan and Eura- wan had presumably related to different source areas in sian continent (Hsieh et al. 2006; Liew and Hsieh 2000) Asia. The second is that there had been probably several allowed immigration of plants from neighboring regions migration routes for plants to migrate from different to Taiwan (Chiang and Schaal 2006). The proportion of source areas to Taiwan. endemic vascular plant species of Taiwan is 26.1%, while This study attempts to identify the source areas of more than 73% of vascular plants in Taiwan are not non-endemic angiosperm species in Taiwan (NEAST). endemic (Hsieh 2002). Non-endemic plant species might Floristic similarities between Taiwan and neighboring have migrated to Taiwan from the neighboring regions, terrestrial ecoregions in Asia were analyzed to iden- or the source area. Nonetheless, that where had been tify the possible source areas. Geographical distribution the source areas of plants in Taiwan is an unanswered ranges of the NEASTs in Asia were explored to mirror question. the possible migration routes of angiosperms from differ- Since few decades ago, tens of botanists had inter- ent source areas to Taiwan. ested on the source areas of angiosperm species in Tai- wan. Several studies had addressed on the phylogenetic Methods relationships of plants among Taiwan and neighboring Data collection and preparation regions, including , Ryukyu archipelago and China, Large quantities of angiosperm data records were utilized to evaluate the immigration of plants from neighboring to analyze floristic relationships and to evaluate geo- regions to Taiwan (Chen et al. 2009; Chiang and Schaal graphical distribution ranges of plants. The angiosperm 2006; Huang et al. 2002; Kokubugata et al. 2011; Wei data records of all the countries in East and South Asia et al. 2010). Some studies had focused on the floristic were downloaded from the Global Biodiversity Informa- compositions to explore the species richness of Taiwan tion Facility (GBIF, http://www.GBIF.org). More than 4.5 and floristic relationships among Taiwan and neighbor- millions data records were collected. The data records of ing regions (Chao et al. 2010; Feroz and Hagihara 2008; angiosperms were mapped by Diva-GIS (Hijmans et al. Hsu and Wolf 2009; Liao et al. 2013; Tang et al. 2013). 2001). Data records with incorrect values of latitude and All the phylogenetic and floristic studies attempted to longitude or without scientific names were dropped out answer two questions: where were the source areas of from the analysis. flora of Taiwan and how many plants had migrated from tropical, subtropical or temperate regions to Taiwan, Angiosperm list of Taiwan respectively. The questions remain unanswered because Angiosperm list of Taiwan is the first and the most continuous vegetation coverage extends from Malay Pen- important dataset of this study. There are two avail- insula northward to the Arctic of Siberia and floristic able angiosperm lists of Taiwan from two different data compositions of vegetations change from tropical to tem- sources. The first list (List I) was derived from the Flora perate regions (Fang et al. 2002; Ni 2003; Ohsawa 1993, of Taiwan, including 3420 angiosperm species (Huang 2006; Olson et al. 2001). The question which vegetation 1996). The second list (List II) was derived from the plant types or which regions are the important source areas of data records downloaded from the GBIF and the number insular non-endemic plant species in Taiwan have never of data records was 560,000. There are more than 8000 been investigated. scientific names in the List II. The numbers of angio- Taiwan is characterized by the massif of a central sperm species are different between List I and List II mountain system with the highest peak of ca. 4000 m because some synonyms are included in the two lists. The above sea level (ASL) and Tropic of Cancer crosses the corrections and revisions of the scientific names from southern part of the island. The northern and southern GBIF and incorporation of the two lists were performed. part of Taiwan belongs to the subtropical and tropical cli- Most of the taxa have the same scientific name in mate zones, respectively. Climates change from southern both List I and List II, whereas some taxa have different Liao and Chen Bot Stud (2017) 58:15 Page 3 of 17

scientific names between List I and List II. If so, the sci- 24°13′ latitude. The N, C and S latitude zones were fur- entific names from GBIF (List II) were recognized as ther divided into eastern (E) and western slopes (W). The accepted names in this study, while that from Flora of border between E and W of the three latitude zones is the Taiwan (List I) synonyms. The scientific names of the mountain ridge of central mountain system, which runs List I and List II were checked on the website of GBIF to from north to south of the island. Seven geographical identify whether the names were accepted names or syn- areas, namely northeast (NE), northwest (NW), central onyms. Synonyms were substituted by accepted names east (CE), central west (CW), southeast (SE), southwest in GBIF and accepted names were reserved in the two (SW), and ST, were obtained in our study. Next, each lists. The accepted names both occurred in the List I and geographical area was divided into several geographi- List II were collected and incorporated into List III. As a cal localities. The NE, NW, CE and CW were divided result, a total of 3275 angiosperm species were confirmed into 3 geographical localities below 2500 m ASL and the in the List III (see Additional file 1: Appendix S1). altitudinal ranges of geographical localities are 0–500 m ASL (lowland and foothill), 500–1500 m ASL (below the Geographical localities of Taiwan cloud zone), and 1500–2500 m ASL (cloud zone). Above The spatial distributions of plants in Taiwan are strongly 2500 m ASL (high elevation above the cloud zone), only affected by climate, elevation, topography, and mon- one geographical locality was obtained for each of N and soon winds (Chao et al. 2010; Chen et al. 1997; Chiou C latitude zones. Division of the S latitude zone is differ- et al. 2010; Hsieh et al. 1997, 1998; Li 2014; Li et al. 2013; ent because the highest elevation is relatively lower. Each Mabry et al. 1998; Su 1984, 1985; Zhang et al. 2010). Tai- of SE and SW was divided into 2 geographical localities wan had been divided into six geographical areas in terms below 1500 m ASL and the altitudinal ranges of geo- of diverse climatic environments and floristic composi- graphical localities are 0–500 and 500–1500 m ASL. tions (Su 1984, 1985). Based on the six geographical areas From 1500 to 3500 m ASL, 2 geographical localities, of Su (1984), we divided Taiwan into several geographi- 1500–2500 and 2500–3500 m ASL, were obtained for cal localities (Fig. 1). First, Taiwan was divided into four the S latitude zone. The ST zone was divided into 2 geo- latitude zones. The southern tip (ST) is at south of 22°13′ graphical localities because elevation of the highest peak latitude and included Hengchun peninsula and Lanyu. is lower than 1000 m ASL. Based on this method, Taiwan The latitude of southern zone (S) is between 22°13′ and was divided into 22 geographical localities. The 560,000 23°13′ latitude, the central zone (C) is between 23°13′ and data records of angiosperms from GBIF were divided 24°13′ latitude, and the northern zone (N) is at north of into 22 data files of 22 geographical localities of Taiwan

Fig. 1 The numbers and acronyms of 22 geographical localities of Taiwan (GLTs). The solid curve represents the profile of the island. Taiwan was divided into four latitude zones, northern zone (N), central zone (C), southern zone (S), and southern tip (ST). The N, C and S latitude zones were divided into eastern (E) and western (W) slopes of central mountain system. Seven geographical areas were obtained, they were NE, NW, CE, CW, SE, SW and ST. The seven geographical areas were then divided into several geographical localities according to elevation. The GLTs were numbered from north to south and from low to high elevations Liao and Chen Bot Stud (2017) 58:15 Page 4 of 17

(GLTs) according to the latitude, longitude and elevations Instead of the aforementioned regions, terrestrial ecore- of data records. The 22 data files were utilized to generate gions proposed by Wild Fund (WWF) (Olson 22 angiosperm lists of 22 GLTs. et al. 2001) were utilized for the floristic analysis (Fig. 2). There are 867 terrestrial ecoregions, classified into 14 dif- Terrestrial ecoregions in Asia ferent biomes on earth. The descriptions of the terrestrial The source areas in relation to the angiosperm species ecoregions can be checked on the website (http://wwf. richness of Taiwan range from tropical islands at equa- panda.org/about_our_earth/ecoregions/ecoregion_list/) tor to subarctic Siberia and extend westward to western and the map of the delineations of the ecoregions was slope of Qinghai–Tibet Plateau. The source areas might downloaded from the internet (http://www.worldwildlife. have included continental Asia, India, Indochina, tropical org/publications/terrestrial-ecoregions-of-the-world). It islands and temperate islands in East Asia. However, the is because of that the ecoregions more accurately reflect aforementioned regions are inappropriate to be used as the complex distribution of the Earth’s natural communi- analyzing units of floristic relationships. Latitude ranges ties (Olson et al. 2001). and area sizes of the aforementioned regions or countries According to the published division of WWF, 38 ter- are wide enough to encompass distinct vegetation types. restrial ecoregions in Asia are included in our study

Fig. 2 Map shows the locations of 34 ecoregions in this study. The acronyms of ecoregions are the same as in Table 1. Floristic relationships among Taiwan and 34 ecoregions were analyzed in this research. The arrows show the five hypothetical migration routes of angiosperms from tropical island (Route I), tropical continent (Route II), subtropical regions (Route III), temperate regions (Route IV) and temperate islands (Route V) to Taiwan (see “Discussion”) Liao and Chen Bot Stud (2017) 58:15 Page 5 of 17

(Olson et al. 2001). However, local studies are not con- Data analysis sistent with the division of the ecoregions. The northern The 34 lists of NEAST of ecoregions were used to evalu- and central China was categorized as one ecoregion in ate geographical distributions of angiosperms and to ana- WWF, while some local references divided northern and lyze floristic similarities among Taiwan and ecoregions. central China into two floristic regions (Fang et al. 2002; The NEASTs were transformed into presence data in Qiu et al. 2011). The Qinling Mountains–Huai River ecoregions (see Additional file 1: Appendix S1). The pres- line (at ca. 34°N) demarcates the northern and central ence data of NEASTs were used for the analysis in this China (Qiu et al. 2011). Therefore, two ecoregions were study to minimize the bias of data information (Beck divided by the Qinlin Mountains-Huai River line in this et al. 2014; Yesson et al. 2007). NEASTs were evaluated study; the northern China was categorized as temper- their geographical distribution patterns in Asia. Geo- ate ecoregion (No. 13) and the central China subtropical graphical distribution range of a NEAST was represented ecoregions (No. 14). In addition, there are 6 ecoregions by the collection of ecoregions that a NEAST is pre- in according to WWF. The 6 ecore- sent. The ecoregions in Asia were categorized into three gions in Indian subcontinent was incorporated into one groups, tropical, subtropical and temperate, according ecoregion because of two reasons. Floristic relationships to the published description of ecoregions in WWF. The between Taiwan and India were supposed to be lower tropical, subtropical and temperate ecoregions were fur- and the number of plant data records at Indian subconti- ther subdivided into several sub-regions in this study. The nent is relatively lower. The ecoregion in Indian subcon- tropical ecoregions were divided into two sub-regions tinent was numbered as 22. Finally, 34 ecoregions were that were ecoregions at tropical continents and tropical obtained (Table 1; Fig. 2). Moreover, the 34 ecoregions islands. The subtropical ecoregions were divided into were classified into four groups in terms of locations and four sub-regions that were southern slopes of Himalaya, climate characteristics. The four groups are cold-winter southwest China, southeast China, and southern islands desert, temperate ecoregions, subtropical ecoregions, of Japan. The temperate ecoregions were divided into and tropical ecoregions (Table 1). Qinghai-Tibet Plateau and Takla-Makan-Gobi Desert, and Siberia, northeast China, peninsula, Angiosperm lists of terrestrial ecoregions and northern islands of Japan. Whether the distribution The angiosperm data records of GBIF were downloaded ranges of NEASTs over these sub-regions or not are the by the countries in Asia. The countries in continental Asia division criterions of distribution types. include Russia, Mongolia, Korea, China, India, countries Floristic similarity between islands and source areas in Indochina, and Malaysia. The temperate islands belong were measured in this study because floristic similarity to Japan and tropical islands include Philippine, Sumatra, provides evidences of dispersal routes from source areas Java, Borneo, Sulawesi, and New Guinea. More than 4.5 to islands (Alsos et al. 2015). In order to analyze the flo- millions of angiosperm data records were downloaded ristic similarities between GLTs and ERGs, the equation from the GBIF. The boundaries of countries in Asia are c/(c + b) was used to calculate the proportion of species different from that of ecoregions. In order to make angi- which given GLT shares with ERG. Within the equation, osperm lists of ecoregions, angiosperm data records c is the number of shared species and b is the number of within an ecoregion were extracted from different coun- species only in given GLT and not in ERG. This measure tries. The angiosperm records extracted from different considers the situation that particular ERG may poten- countries were manipulated by ArcMap software (ESRI, tially contain other species which are not in given GLT Redlands, USA). The extracted data records were incor- and which does not necessarily to influence the similarity porated into an ecoregion. The incorporated data records of that GLT and ERG. The angiosperm lists of 22 GLTs of ecoregions were mapped, checked and revised by and 34 ecoregions of Asia were combined to make a data Diva-GIS (Hijmans et al. 2001) and were utilized to gen- matrix for the cluster analysis. Simple nearest neighbor- erate angiosperm lists of ecoregions. Numbers of data ing algorithm and Bray–Curtis method were chosen for records and numbers of angiosperm lists of ecoregions clustering. The computations were conducted by means are listed in Table 1. of PC-ORD program (v.5) (McCune and Mefford 2006). The angiosperm lists of ecoregions were compared with that of List III (angiosperm list of Taiwan). Non- Results endemic angiosperm species in Taiwan (NEAST) were Numbers of NEAST in ecoregions of Asia evaluated their geographical distribution ranges in Asia. The numbers of family, genus and species of NEAST in The NEASTs of each ecoregion were collected to gener- ecoregions are showed in Table 2. Four subtropical ecore- ate angiosperm lists and, eventually, there are 34 lists of gions, No. 8, 14, 17 and 20, and one tropical ecoregions, NEASTs of ecoregions in this study. No. 26, possess high numbers of NEAST (Table 2). Three Liao and Chen Bot Stud (2017) 58:15 Page 6 of 17

Table 1 The names and abbreviations of biomes and 34 terrestrial ecoregions in Asia Acronym of Name of biome Name of Number of Number of ecoregions ecoregions data recorda scientific nameb

Group of cold winter deserts ERG-15 Cold-winter desert Takla-Makan-Gobi Desert 12,062 2744 ERG-16 Cold-winter desert Tibetan 38,004 2371 Group of temperate ecoregions ERG-02 Mixed mountain system Altai highlands 2355 162 ERG-03 Temperate grasslands Mongolian-Manchurian Steppe 2071 600 ERG-01 Temperate needle-leaf forests/woodlands East Siberian Taiga 1343 467 ERG-04 Temperate broad-leaf forests Manchu-Japanese mixed forests 3180 909 ERG-05 Temperate broad-leaf forests Manchu-Japanese mixed forests 5575 988 ERG-06 Temperate broad-leaf forests Oriental deciduous forests 442 238 ERG-07 Temperate broad-leaf forests Oriental deciduous forests 33,562 2066 ERG-11 Temperate broad-leaf forests Oriental deciduous forests 3618 1085 ERG-13 Temperate broad-leaf forests Oriental deciduous forests 13,876 2959 Group of subtropical ecoregions ERG-08 Subtropical/temperate rain forests/woodlands Japanese evergreen forest 287,306 4433 ERG-09 Subtropical/temperate rain forests/woodlands Japanese evergreen forest 22,180 1857 ERG-10 Subtropical/temperate rain forests/woodlands Japanese evergreen forest 1684 611 ERG-12 Subtropical/temperate rain forests/woodlands Japanese evergreen forest 600,788 4519 ERG-14c Temperate broad-leaf forests Oriental deciduous forests 139,697 7862 ERG-17 Subtropical/temperate rain forests/woodlands Chinese subtropical forests 194,854 11,696 ERG-20 Mixed mountain system Szechwan highlands 186,933 12,271 ERG-21 Mixed mountain system Himalayan highlands 55,145 6413 Group of tropical ecoregions ERG-18 Tropical humid forests South Chinese rainforests 9396 2586 ERG-19 Tropical humid forests South Chinese rainforests 5770 1876 ERG-23 Tropical humid forests Bengalian rainforest 2230 412 ERG-25 Tropical humid forests Burman rainforest 992 439 ERG-27 Tropical humid forests Indochinese rainforests 196,220 5577 ERG-28 Tropical humid forests Malayan rainforests 14,573 3339 ERG-22 Tropical dry forest/woodlands India subcontinent 28,662 3112 ERG-24 Tropical dry forests Burma monsoon forests 2075 573 ERG-26 Tropical dry forests/woodlands Thailandian monsoon forest 33,293 5996 ERG-29 Mixed island systems Philippines 33,825 3390 ERG-30 Mixed island systems Borneo 90,332 6813 ERG-31 Mixed island systems Sumantra 11,763 2341 ERG-32 Mixed island systems Java 7431 1503 ERG-33 Mixed island systems Sulawesi 20,724 2391 ERG-34 Mixed island systems Papuan 226,960 9502

Locations of the ecoregions are showed on the map in Fig. 1. ERG is the abbreviation of ecoregion. The 34 ecoregions were categorized into 4 groups. Numbers of data records were downloaded from the Global Biodiversity Information Facility (GBIF) and numbers of scientific names were derived from the data records a Number of data record: data records were downloaded from the GBIF and data records of an ecoregion were extracted from different countries and incorporated into one ecoregion b Number of scientific name: the number of scientific names derived from the data records of GBIF. The number is not the real species richness because many synonyms were in the lists c ERG-14 is categorized into subtropical group according to local references and different from the categorization of WWF Liao and Chen Bot Stud (2017) 58:15 Page 7 of 17

Table 2 The numbers of family, genus, and species of NEAST than tropical and subtropical ecoregions dem- of ecoregions in Asia listed in the table were the numbers onstrated lower floristic relationships. On the other hand, of non-endemic angiosperm species in Taiwan (NEAST) Philippine is the closest tropical island to Taiwan in dis- Acronym of ecoregions Family Genus Species tance, but there are only 463 NEASTs in Philippine. The Papua New Guinea is a tropical island distant from Tai- Cold winter desert wan, while the number of NEAST of Papua New Guinea ERG-15 65 168 222 is higher than that of Philippine. In contrast to tropical ERG-16 28 93 117 islands, angiosperm species richness of subtropical and Temperate regions tropical ecoregions at Asian continent are more closely ERG-02 4 13 14 related to Taiwan. ERG-03 14 37 43 ERG-01 16 30 33 Geographical distribution types of NEAST in Asia ERG-04 37 94 115 Geographical distribution ranges of 3275 angiosperm ERG-05 55 133 170 species were examined in terms of the current occur- ERG-06 29 39 46 rences of angiosperm data records (see Additional file 1: ERG-07 101 337 495 Appendix S1). Of the 3275 angiosperms, 847 species are ERG-11 54 155 190 endemic in Taiwan and 2428 are NEASTs. The geograph- ERG-13 102 315 456 ical distribution ranges of the 2428 NEASTs were catego- Subtropical regions rized into 25 types belong to 7 groups (Fig. 3; Table 3). ERG-08 140 575 1003 NEASTs of Group B-II has their distribution ranges ERG-09 120 439 699 from tropical islands to Taiwan and these plants are absent ERG-10 82 173 226 at Asian continent. Meanwhile, Group G has their distri- ERG-12 126 488 791 bution ranges only from temperate ecoregions to Taiwan. ERG-14 155 662 1198 Distribution ranges of Group B-II and Group G are absent ERG-17 161 841 1631 at tropical and subtropical ecoregions at Asian continent ERG-20 149 742 1344 and are obviously different from most of the NEASTs. In ERG-21 104 462 734 contrast to these two groups, more than 88% of all the Tropical regions NEASTs have their distribution ranges over tropical, sub- ERG-18 136 547 886 tropical or temperate ecoregions at Asian continent. ERG-19 109 381 551 ERG-23 33 76 96 Cluster analysis and floristic similarities among GLTs ERG-25 33 90 112 and ecoregions ERG-27 131 546 917 Floristic similarities of pairs between 22 GLTs and 34 ERG-28 96 292 423 ecoregions in Asia are listed in Table 4. Floristic similari- ERG-22 89 315 486 ties are evidently low between all the GLTs and ecore- ERG-24 37 85 108 gions at higher latitudes. The values of similarity indices ERG-26 140 600 1007 between all the GLTs and ecoregions at higher latitudes, ERG-29 104 377 541 from ERG-01 to ERG-06, are less than 0.1. Higher flo- ERG-30 100 332 469 ristic similarities are observed between some GLTs and ERG-31 75 195 256 5 subtropical and 2 tropical ecoregions, they are No. 08, ERG-32 68 202 274 14, 17, 18, and 20. The second higher floristic similarities ERG-33 96 314 435 are observed between some GLTs, especially the GLTs ERG-34 121 496 827 at lower elevations, and tropical ecoregions, No. 26, 27, The numbers after ERG and dash are the numbers of ecoregions listed in Table 1 and 34. Interestingly, similarities indices between GLTs at and Fig. 2 higher elevations and subtropical ecoregions are higher than that between GLTs at higher elevations and temper- ate ecoregions. of the four subtropical ecoregions, No. 14, 17, and 20, are Floristic similarities among GLTs and subtropical and the ecoregions at southern China, while the other one at tropical ecoregions in Asia are supported by the cluster Japan. The tropical ecoregion with high number of NEAST analysis. Simple nearest neighboring algorithm resulted 2 is at Indochina. Subtropical ecoregions at southern China major clusters (Fig. 4). Temperate and cold-winter desert has evidently high floristic relationships with Taiwan. ecoregions at Asian continent were came out as the first Temperate ecoregions possess relatively lower numbers cluster and were the most deviating ecoregions. Many Liao and Chen Bot Stud (2017) 58:15 Page 8 of 17

Fig. 3 The maps show the 25 distribution types of 7 distribution Groups of non-endemic angiosperm species in Taiwan (NEAST). Angiosperms of Group A distribute from tropical islands to subarctic Siberia. B-I to B-IV are tropical distribution. C-I to C-III are tropical to subtropical distribution. D-I to D-VIII are tropical to temperate distribution. E-I to E-II are subtropical distribution. F-I to F-VI are subtropical to temperate distribution. G is temperate distribution. There are 2428 NEASTs evaluated in this research. The detail descriptions of distribution ranges and the numbers of species are listed in Table 3 Liao and Chen Bot Stud (2017) 58:15 Page 9 of 17

Table 3 Geographical distribution groups of non-endemic angiosperm species in Taiwan (NEAST) Distribution group Type Distribution range Number of species

Widely distribution (Group A) A-I Widely distribution from tropical through subtropical to temperate 27 ecoregions or to cold winter deserts Tropical distribution (Group B) B-I Tropical islands, Indochina, India, and few species reach Japan 156 B-II Tropical islands 138 B-III India and Indochina 27 B-IV Indochina, Hainan, southeast coast of China 24 Tropical to subtropical distribution (Group C) C-I Tropical islands, Indochina, India, southern slope of Qin-zang Plateau; 398 south, central and north China C-II Tropical islands, southern slope of Qing-zang Plateau and south, central 33 and north China C-III Indochina, southern slope of Qing-zang Plateau, south, central and 232 north China Tropical to temperate distribution (Group D) D-I Tropical islands, Indochina, India, southern slope of Qing-zang Plateau; 74 south, central and north China, Japan, Korea, northeast China D-II Tropical islands, Indochina, India, southern slope of Qing-zang Plateau; 25 south, central and north China, Japan, Korea, northeast China, Mon‑ golia, Altai D-III From Indochina and India toward north to Siberia 17 D-IV Indochina, India, southern slope of Qing-zang Plateau, Xinjiang, south, 57 central, north and northeast China, Japan, Korea D-V Tropical islands, Indochina, India, southern slope of Qing-zang Plateau, 233 south, central and north China, Korea and Japan D-VI Tropical islands, southern slope of Qing-zang Plateau, south, central and 46 north China, Korea and Japan D-VII Indochina, India, southern slope of Qing-zang Plateau, south, central 84 and north China, Korea and Japan D-VIII Indochina, southern slope of Qing-zang Plateau, south, central and 55 north China, Korea and Japan Subtropical distribution (Group E) E-I Hainan, southeast coast of China, southeast, central and north China 101 E-II Hainan, southeast coast of China, southern slope of Qing-zang Plateau, 164 south, central and north China, Korea Subtropical to temperate distribution (Group F) F-I Southern slopes of Qing-zang Plateau, Tibet, Xinjiang, south, central, 26 north, and northeast China, Japan, Korea and Siberia F-II Hainan, southeast coast of China, Southern slopes of Qing-zang Plateau, 50 Tibet, Xinjiang, south, central, north, and northeast China, Japan, Korea and Siberia F-III Southern slopes of Qing-zang Plateau, south, central, and north China 35 Japan and Korea F-IV Hainan, southeast coast of China, Southern slopes of Qing-zang Plateau, 30 Tibet, Xinjiang, south, central, north, and northeast China, and Korea F-V southeast coast of China, Southern slopes of Qing-zang Plateau, south, 152 central, and north China Japan and Korea F-VI Hainan, southeast coast of China, south, central, and north China, Japan 94 Temperate distribution (Group G) G-I Northeast China, Japan and Korea, few species reach Hainan, southeast 150 coast of China and tropical islands

The NEASTs were categorized into 7 distribution groups and 25 types according to their distribution ranges in Asia. Descriptions of geographical distribution ranges and numbers of angiosperm species of each type were listed in the table tropical, subtropical ecoregions and GLTs formed the ecoregions. However, the assumption is not supported by second cluster. Deep in the second cluster, three subtrop- our analysis. The results of floristic similarity and cluster- ical ecoregions, two tropical ecoregions, and many GLTs ing analysis both concluded floristic dissimilarity among were grouped together. GLTs at higher elevations and ecoregions at higher lati- The GLTs at northern Taiwan and high elevations were tudes in Asia continent. Although some angiosperms in assumed to have floristic similarities with temperate GLTs at higher elevation are observed in higher latitudes, Liao and Chen Bot Stud (2017) 58:15 Page 10 of 17 0.20 0.24 0.22 0.23 0.22 0.16 0.23 0.23 0.26 0.23 0.24 0.23 0.25 0.31 0.28 0.31 0.29 0.29 0.15 0.29 0.21 0.21 ERG-21 0.43 0.48 0.47 0.47 0.45 0.39 0.47 0.50 0.50 0.49 0.51 0.47 0.44 0.54 0.51 0.53 0.52 0.52 0.50 0.42 0.42 0.33 ERG-20 0.55 0.58 0.56 0.58 0.55 0.50 0.58 0.60 0.61 0.58 0.63 0.59 0.54 0.63 0.60 0.63 0.65 0.65 0.37 0.65 0.49 0.49 ERG-17 0.43 0.49 0.50 0.52 0.49 0.42 0.45 0.46 0.48 0.51 0.47 0.32 0.36 0.43 0.42 0.45 0.49 0.51 0.49 0.45 0.46 0.34 ERG-14 0.21 0.28 0.27 0.31 0.29 0.12 0.20 0.23 0.27 0.25 0.26 0.28 0.21 0.26 0.22 0.27 0.26 0.32 0.19 0.28 0.28 0.28 ERG-12 0.09 0.09 0.10 0.11 0.10 0.08 0.08 0.09 0.09 0.10 0.11 0.09 0.08 0.08 0.07 0.08 0.09 0.10 0.07 0.08 0.10 0.08 ERG-10 0.22 0.27 0.28 0.32 0.28 0.14 0.20 0.23 0.26 0.26 0.28 0.27 0.20 0.24 0.22 0.26 0.27 0.31 0.21 0.27 0.26 0.25 ERG-09 0.35 0.36 0.40 0.37 0.29 0.36 0.36 0.36 0.21 0.29 0.31 0.36 0.34 0.36 0.41 0.36 0.29 0.34 0.30 0.26 0.34 0.34 ERG-08 Group of subtropical ecoregions of subtropical Group 0.13 0.18 0.20 0.20 0.18 0.06 0.14 0.14 0.17 0.18 0.18 0.16 0.12 0.16 0.15 0.18 0.17 0.19 0.12 0.16 0.19 0.19 ERG-13 0.04 0.07 0.06 0.07 0.07 0.03 0.05 0.05 0.07 0.06 0.06 0.06 0.05 0.07 0.05 0.07 0.06 0.08 0.04 0.09 0.08 0.08 ERG-11 0.15 0.21 0.20 0.24 0.22 0.05 0.12 0.13 0.18 0.16 0.18 0.18 0.11 0.14 0.12 0.15 0.15 0.19 0.15 0.18 0.21 0.22 ERG-07 0.02 0.02 0.02 0.03 0.03 0.00 0.01 0.01 0.02 0.01 0.01 0.02 0.01 0.01 0.01 0.01 0.01 0.02 0.01 0.02 0.03 0.03 ERG-06 0.04 0.08 0.07 0.08 0.08 0.00 0.03 0.03 0.05 0.05 0.05 0.06 0.03 0.04 0.03 0.03 0.04 0.06 0.07 0.04 0.09 0.09 ERG-05 0.03 0.05 0.05 0.06 0.05 0.01 0.03 0.03 0.04 0.04 0.04 0.04 0.02 0.04 0.03 0.04 0.04 0.05 0.03 0.05 0.05 0.06 ERG-04 0.01 0.02 0.01 0.02 0.02 0.00 0.01 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.02 0.01 0.02 0.01 0.02 0.02 0.02 ERG-03 0.00 0.01 0.01 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.01 0.01 ERG-02 0.01 0.01 0.01 0.01 0.01 0.00 0.01 0.00 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.01 0.00 0.02 0.01 0.01 ERG-01 Group of temperate ecoregions of temperate Group 0.02 0.05 0.04 0.05 0.05 0.01 0.02 0.03 0.04 0.03 0.03 0.03 0.03 0.04 0.03 0.04 0.03 0.04 0.03 0.05 0.05 ERG-16 0.06 0.07 0.09 0.09 0.10 0.10 0.04 0.06 0.06 0.07 0.07 0.08 0.07 0.05 0.07 0.07 0.07 0.08 0.08 0.04 0.08 0.10 ERG-15 Cold winter desert winter Cold 0.10 Floristic similarities between geographical localities of Taiwan (GLT) and 34 terrestrial ecoregions (ERG) in Asia (ERG) ecoregions and 34 terrestrial (GLT) localities of Taiwan similarities between geographical Floristic - - 2500 - 1500 - 2500 - 1500 - 500 - 1500 - 500 - 1000 - 500 - 3500 - 500 19. S - 2500 18. MW 17. ME - 2500 16. NW 15. NE - 2500 14. ST 13. SW 12. SE - 1500 11. MW 10. ME - 1500 9. NW 8. NE - 1500 7. ST 6. SW 5. SE - 500 4. MW 3. ME - 500 2. NW 22. S - 3500 1. NE - 500 21. M 4 Table Number of ecore gion 20. N - 3500 Liao and Chen Bot Stud (2017) 58:15 Page 11 of 17 0.38 0.37 0.41 0.36 ERG-34 0.33 0.27 0.34 0.21 0.24 0.27 0.21 0.22 0.25 0.24 0.16 0.15 0.19 0.15 0.17 0.15 0.12 0.07 ERG-33 0.17 0.14 0.19 0.20 0.21 0.22 0.21 0.13 0.13 0.16 0.13 0.16 0.15 0.16 0.08 0.07 0.10 0.08 0.11 0.06 0.04 0.02 ERG-32 0.11 0.09 0.12 0.13 0.13 0.14 0.14 0.08 0.08 0.09 0.08 0.11 0.10 0.10 0.05 0.05 0.07 0.04 0.07 0.05 0.04 0.02 ERG-31 0.09 0.06 0.12 0.11 0.12 0.11 0.13 0.08 0.08 0.09 0.08 0.11 0.10 0.11 0.06 0.06 0.07 0.06 0.08 0.05 0.04 0.04 ERG-30 0.18 0.17 0.21 0.22 0.23 0.23 0.23 0.15 0.15 0.17 0.14 0.17 0.17 0.16 0.09 0.09 0.12 0.09 0.12 0.07 0.05 0.04 ERG-29 0.20 0.15 0.22 0.23 0.26 0.25 0.25 0.16 0.16 0.18 0.16 0.20 0.19 0.19 0.11 0.11 0.13 0.13 0.16 0.11 0.09 0.07 ERG-28 0.17 0.13 0.16 0.19 0.19 0.22 0.20 0.11 0.12 0.14 0.11 0.16 0.14 0.14 0.07 0.07 0.09 0.06 0.10 0.05 0.04 0.03 0.36 0.37 0.42 0.39 0.45 0.41 ERG-27 0.33 0.30 0.29 0.31 0.27 0.34 0.34 0.33 0.19 0.18 0.23 0.18 0.26 0.15 0.12 0.13 ERG-26 0.41 0.40 0.43 0.45 0.41 0.47 0.44 0.37 0.35 0.38 0.34 0.35 0.39 0.39 0.26 0.25 0.30 0.27 0.31 0.21 0.18 0.20 ERG-25 0.05 0.03 0.05 0.06 0.05 0.06 0.05 0.04 0.03 0.04 0.03 0.03 0.04 0.04 0.02 0.02 0.03 0.03 0.03 0.02 0.02 0.02 ERG-24 0.05 0.07 0.05 0.06 0.05 0.06 0.06 0.04 0.04 0.05 0.04 0.04 0.05 0.04 0.03 0.03 0.03 0.04 0.04 0.03 0.01 0.01 ERG-23 0.04 0.04 0.04 0.05 0.05 0.06 0.04 0.03 0.03 0.03 0.02 0.03 0.03 0.03 0.02 0.02 0.02 0.03 0.02 0.02 0.01 0.00 ERG-22 0.20 0.17 0.20 0.24 0.21 0.26 0.21 0.13 0.13 0.17 0.12 0.15 0.15 0.14 0.09 0.09 0.12 0.11 0.11 0.08 0.07 0.05 ERG-19 0.24 0.21 0.27 0.29 0.27 0.29 0.30 0.23 0.21 0.21 0.19 0.29 0.24 0.24 0.14 0.14 0.17 0.14 0.19 0.12 0.08 0.08 Group of tropical ecoregions of tropical Group ERG-18 0.40 0.38 0.41 0.44 0.38 0.45 0.43 0.35 0.33 0.35 0.31 0.37 0.36 0.36 0.25 0.24 0.29 0.24 0.30 0.19 0.17 0.13 continued - 1500 - 2500 - 2500 - 1500 - 500 - 500 - 1500 - 1000 - 500 - 3500 - 500 4 Table Number of ecoregion of the abbreviation names are row The 2 . as in Fig. the ecoregions are columns The GLT. and ERG and b is the species only in given GLT given between c is the species shares c/(c + b), equation by calculated was index The of similarity the values higher than 0.35 indices italics in the table are The 1 . as in Fig. GLTs 1. NE - 500 3. ME - 500 2. NW 4. MW 5. SE - 500 8. NE - 1500 7. ST 6. SW 10. ME - 1500 9. NW 12. SE - 1500 11. MW 15. NE - 2500 14. ST 13. SW 17. ME - 2500 16. NW 19. S - 2500 18. MW 20. N - 3500 22. S - 3500 21. M Liao and Chen Bot Stud (2017) 58:15 Page 12 of 17

Fig. 4 Floristic relationships between 22 geographical localities of Taiwan (GLT) and 34 ecoregions in Asia were analyzed by cluster analysis, using simple nearest neighbor algorithm and Bray–Curtis dissimilarity method. Non-endemic angiosperm species in Taiwan (NEAST) were used for the cluster analysis. The ecoregions of cold winter desert and temperate needle-leaved forests were the first cluster and came out as the most deviating regions. The second cluster was comprised of many ecoregions and GLTs. Deep in the second cluster, tropical and subtropical ecoregions and some GLTs were grouped together. ERG is the acronym of ecoregion. The numbers after ERG and dash are the number of ecoregions listed in Table 1 and Fig. 2. The number and acronyms of 22 GLTs are the same as in Fig. 1 Liao and Chen Bot Stud (2017) 58:15 Page 13 of 17

they are not dominant members. A further examina- tropical and subtropical geographical distribution ranges. tion on the floristic lists of GLTs demonstrated that In other words, NEASTs with tropical and subtropical angiosperms with tropical and subtropical distributions distributions extend their ranges from low to high eleva- (Group B, C, D, E, F) extend their distribution ranges tions in Taiwan that have caused low floristic similarities from low to high elevations in Taiwan (see later texts). among GLTs at high elevations and ecoregions at higher This is most likely the explanation of the low floristic latitudes (Fig. 4; Table 4). Results from geographical dis- similarity between Taiwan and temperate ecoregions in tribution range, cluster analysis and similarity indices all Asian continent. concluded floristic similarities among GLTs and tropical and subtropical ecoregions in Asian continent. Proportions of geographical distribution types of NEAST in Taiwan Discussion NEASTs with tropical and subtropical geographical dis- Recently, studies on large scale biodiversity pattern are tribution ranges (Group B, C, D, E, and F) are the domi- available because of rapid accumulation of global data nant members of all the GLTs (Fig. 5). The 7 geographical records (Beck et al. 2014; Flemons et al. 2007; Gural- distribution groups exhibit similar ratio among GLTs, nick et al. 2007; Ingwersen and Chavan 2011; Saarenmaa though slightly differences has been observed. The pro- 2005). GBIF is the largest online provider of global data portions of Group B, C and D are relatively higher in records (Flemons et al. 2007; Padial et al. 2010). Large southern GLTs at S and ST latitude zones, while that of quantities of biodiversity data are inevitably spatially Group E and F in northern GLTs at N latitude zone and biased due to uneven effort of data collections and it is higher elevations. The southern GLTs at S and ST latitude problematic to use species occurrence data in species dis- zones possess relatively more NEASTs with tropical dis- tribution models (Beck et al. 2014). Subsampling distri- tributions, whereas the northern GLTs at N latitude zone bution data had been used to reduce spatial bias of data more NEASTs with subtropical distributions. and improved model quality (Beck et al. 2014). Our study The GLTs above 3000 m ASL were expected to be attempts to minimize the bias error of species occur- dominated by NEASTs with temperate distributions rence data by transforming georeferenced data into pres- (Group G); however, they are dominated by NEASTs with ence-absence data. Our results of floristic relationships

Fig. 5 The proportions of angiosperms with seven geographical distribution groups show different trends among geographical localities of Taiwan (GLTs). Angiosperms with tropical and subtropical distribution are dominant members in all the GLTs. The proportions of angiosperms with tropical distributions (Group B, C and D) increase from northern to southern Taiwan but decrease from low to high elevations. The proportions of angiosperms with subtropical distributions (Group E and F) present opposite trends to that with tropical distributions. Angiosperms with temper‑ ate distribution (Group G) are not dominant members of GLTs at higher elevations. The numbers at X axis are the numbers of GLTs listed in Fig. 1. At bottom of the figure, four lines and four numbers below the lines represent the upper altitudinal limits of GLTs Liao and Chen Bot Stud (2017) 58:15 Page 14 of 17

between GLTs and ecoregions in Asia are acceptable species richness of family Fagaceae is higher at subtropi- because of few reasons. The first is that the numbers of cal regions than at tropical or temperate regions in East scientific names are higher at subtropical ecoregions in Asia (Liao and Chen 2015). The study of Fagaceae species this study. The species richness pattern presented by the richness is similar to our results in Table 1. In summary, numbers of scientific names in Table 1 agreed with the angiosperm species richness along latitudes in Asia is proposed latitudinal pattern of species richness in Asia. worth to be investigated to understand the patterns and Subtropical region in East Asia were proposed to have causal factors of species richness along latitudes in Asia. high species richness along the latitudinal gradient (Feng Latitudinal patterns of angiosperm species richness et al. 2016; López-Pujol et al. 2011; Qiu et al. 2011; Zhu in Asia are important on the angiosperm specie rich- 1997). In this study, higher numbers of scientific names ness in Taiwan because latitudinal range and climatic in two subtropical ecoregions agreed with the proposed environments of Taiwan are similar to the subtropical latitudinal change of species richness, despite the num- regions in continental Asia. However, subtropical region bers of data records are not the highest (Table 1). In is likely not the only source area of insular non-endemic addition, not all the angiosperm species of ecoregions angiosperm species in Taiwan. Some phylogenetic stud- were utilized in this study but only the species that are ies investigated population genetic variations to evaluate common between Taiwan and ecoregions, the NEASTs. historical migration processes of plants from neighboring The utilization of species common between Taiwan and regions to Taiwan and five hypothetical migration routes ecoregions diminished the effects of spatial bias of data have been proposed (Fig. 2) (Huang 2011, 2014; Shen collections and minimized the study error caused by spa- 1997; Wang 1992a, b). The five routes provided path- tial bias of database. ways for plants to migrate from tropical islands (Route I), Interestingly, our study has implied an extraordinary Indochina (Route II), southern China (Route III), north- species richness pattern along latitudes in East Asia. east China (Route IV) and temperate islands (Route V) to A latitudinal pattern of angiosperm species richness Taiwan (Fig. 2). Our study attempts to identify the extent appears while numbers of scientific names are compared to which the five migration routes affected on the angio- among ecoregions of four groups (Table 1). Two subtrop- sperm species richness of Taiwan. ical ecoregions in South China possess the highest num- Our results of floristic similarities and geographical ber of scientific names in Asia and there are 12,271 and distributions both demonstrated that angiosperm spe- 11,696 scientific names of ecoregions 20 and 17, respec- cies richness of Taiwan are closely related to tropical tively (Table 1). High numbers of scientific names in sub- ecoregions at Indochina and subtropical ecoregions at tropical ecoregions are expected to have high angiosperm southern China. The results support that the Route II species richness. and Route III are important on the angiosperm species Generally, species richness is highest at tropical regions richness in Taiwan. The Route II provided pathway for and decreases toward higher latitudes (Barthlott et al. the migration of plants from tropical ecoregions or Indo- 2007; Pianka 1966; Qian et al. 2003, 2007). To our knowl- china through Southern China Sea to Taiwan (Huang edge, higher angiosperm species richness in subtropical 2014; Shen 1997) and angiosperms of Group B, C, and regions than in tropical regions is a novel species rich- D was able to migrate to Taiwan through Route II. The ness pattern along latitudes. Higher species richness at Route III is from the eastern slope of Qinghai-Tibet Pla- subtropical region observed in this study might be false teau through southern China to Taiwan (Matuszak et al. because some factors may lead to a false species richness 2016; Wang 1992a, b). The Route III provided the path- pattern along latitudes. The factors include bias data col- way for the migration of plants from subtropical ecore- lections, synonyms in lists of ecoregions, various sizes of gions to Taiwan and angiosperms of Group C, D, E, and F ecoregions’ areas, etc. Although these factors do not sup- were able to migrate to Taiwan through Route III. port our observation, our observation cannot be rejected The plants of tropical islands had presumably migrated because of some reasons. The first, subtropical region to Taiwan via long distance dispersal and angiosperms had been proposed as one of the biodiversity centers in of B-II probably migrated to Taiwan by using Route I. Asia since several decades ago (Wang 1992a, b). The sub- However, tropical islands in south Asia, include Philip- tropical region had served as a refuge for angiosperms in pine and New Guinea, are not as important as Asian Asia during ice age in the Quarternary (Qian and Rick- continents on the angiosperm species richness of Taiwan lefs 2000; Qian et al. 2003; Qiu et al. 2011; Tiffney 1985). because of two reasons. The first, the number of NEAST These studies had offered evidences supporting high is lower in tropical islands than in Asian continent. The species richness of subtropical regions in Asia. The sec- second, floristic relationships between Taiwan and tropi- ond, a recent study had also implied higher plant species cal islands are not as close as that between Taiwan and richness at subtropical regions. The study proposed that Asian continents. Land bridge connections had never Liao and Chen Bot Stud (2017) 58:15 Page 15 of 17

existed between Philippine and Taiwan in the Quaternary age (Huang and Lin 2006). Migrations of plants from Tai- (Voris 2000; Zeng 1993); therefore, plants migrated from wan to northern latitudes or vice versa were supported tropical islands to Taiwan were probably dispersed by by some evidences of published documents. Analysis on currents, winds or birds. It is likely because of that migra- the population genetics is suggested to provide evidence tion of plants through long distance dispersal is more dif- for identifying the migration of plants from northern ficult than that through land bridge connections. latitudes to Taiwan or vice versa; to date, few study has In contrast to tropical and subtropical ecoregions, tem- focused on this topic. Route IV and V had provided path- perate ecoregions are far less important on angiosperm way for plants to migrate from temperate regions to Tai- species richness in Taiwan. The Route IV and Route V wan, whereas there are only 150 species of Group G. The provided pathways for the migration of plants from tem- number of Group G indicated that Route IV and V are perate ecoregions to Taiwan (Fig. 2). The Route IV is less important on angiosperm species richness in Taiwan. from northeast China through Korea, Yellow Sea, East Our results concluded that tropical and subtropi- China Sea to Taiwan (Shen 1997). The Route V is from cal regions at Asian continent were most important on Japan through Ryukyu archipelago to Taiwan (Huang and the angiosperm species richness in Taiwan because of Lin 2006). The angiosperms of Group D and F extend dis- high floristic similarity between GLTs and tropical and tribution ranges from tropical or subtropical to temper- subtropical ecoregions. Tropical and subtropical ecore- ate ecoregions had probably migrated to Taiwan through gions are most likely the most important source areas Route IV or V. However, it is questionable whether angio- of angiosperms in Taiwan. The tropical islands at South sperms of Group D and F had migrated from southern to Asia are the second important. Angiosperms with dis- northern latitudes or vice versa. A published paper had tribution ranges over tropical and subtropical ecore- proposed several hypothetical migration routes from gions are the dominant members of the vegetation from southern to northern latitudes in China (Wang 1992a, low to high elevations in Taiwan. Most of the angio- b). Angiosperms had migrated from Southwest China sperms at high elevation in Taiwan are the species with toward north, northeast, or east to northern China, tropical and subtropical distributions and, therefore, northeast China, Japan or Taiwan, respectively (Wang temperate ecoregions in Asia are less important on the 1992a, b). Therefore, angiosperms with distribution from angiosperm species richness in Taiwan. Infraspecific tropical or subtropical to temperate ecoregions (Group D genetic variations of angiosperms in Asia is suggested and F) had probably migrated from southern to northern to be investigated to determine the historical migration latitudes or from southern China to Taiwan. The studies of angiosperms from tropical or subtropical regions to of Wang (1992a, b) had implied that most of the plants Taiwan. with distributions from tropical or subtropical to tem- Additional file perate regions had probably migrated from southern to northern latitudes. Therefore, Route IV and V were not Additional file 1: Appendix S1. Presence-absence data of angiopserms supported by the studies of Wang (1992a, b). Although examined in this study. few evidences had indicated migrations of gymno- sperms from temperate regions to southern latitudes (Li et al. 2003), more evidences are necessary for the iden- Abbreviations ASL: above sea level; GBIF: Global Biodiversity Information Fertility; GLT: tification of Route IV and Route V for the migration of geographical locality of Taiwan; NEAST: non-endemic angiosperm species in angiosperms. Taiwan; WWF: World Wild Fund. Angiosperms of Group G have not been observed at Authors’ contributions tropical or subtropical ecoregions in Asian continent. CCL and CHC conceived and designed the survey. CCL and CHC formulated These plants had probably migrated from northern lati- the idea for the paper. CCL performed data analyses and wrote the paper. tudes to Taiwan or vice versa. Migration of plants from Both authors read and approved the final manuscript. northern latitudes to Taiwan was supported by the phy- Author details logenetic studies of Chamaecyparis formosensis and 1 Department of Life Science, Chinese Culture University, 55, Hwa‑Kang Road, Chamaecyparis obtusa var. formosana (Li et al. 2003). Yang‑Ming‑Shan, Taipei 11114, Taiwan, ROC. 2 Endemic Species Research Institute, Nantou County, Taiwan, ROC. Migration of plants from northern to southern latitudes might have been caused by global cooling in the late Acknowledgements Tertiary and Quaternary that had forced many plants to The authors thank two anonymous reviewers for their valuable and critical comments and substantial improvements. We appreciated the works of migrate southward (Chung et al. 2011; Denk 2004; Huang Dr. Min Liu in Department of Life Science, Chinese Culture University. She and Lin 2006; Huang et al. 2004). Meanwhile, phyloge- reviewed and edited the manuscript and conceived new ideas. The authors netic study had identified thatTrochodendron aralioides want to thank Dr. Chang-Fu Hsieh in the Institute of Ecology and Evolutionary Biology, National Taiwan University for providing lists of Flora of Taiwan and Dr. had expanded from Taiwan to northern latitudes after ice Liao and Chen Bot Stud (2017) 58:15 Page 16 of 17

Shong Huang and Pei-Chun Liao in the Department of Life Science, National Fernández-Palacios JM, Nascimento LD, Otto R, Delgado JD, García-del-Rey E, Taiwan Normal University for valuable comments. Mr. Shih-Chieh Kuo in the Arévalo JR, Whittaker RJ (2011) A reconstruction of Palaeo-, Department of Life Science, Chinese Culture University helped us to establish with particular reference to the long-term biogeography of the Atlantic the database and Mr. Kai-Jie Yang in the Institution of Geology, Chinese Cul‑ island laurel forests. J Biogeogr 38:226–246 ture University provided technical support. Feroz SM, Hagihara A (2008) Comparative studies on community ecology of two types of subtropical forests grown in silicate and limestone habitats Competing interests in the northern part of Okinawa Island, Japan. Taiwania 53:134–149 The authors declare that they have no competing interests. Flemons P, Guralnick R, Krieger J, Ranipeta A, Neufeld D (2007) A web-based GIS tool for exploring the world’s biodiversity: the Global Biodiversity Received: 13 July 2016 Accepted: 20 March 2017 Information Facility Mapping and Analysis Portal Application (GBIF- MAPA). Ecol Informatics 2:49–60 Guralnick RP, Hill AW, Lane M (2007) Towards a collaborative, global infrastruc‑ ture for biodiversity assessment. Ecol Lett 10:663–672 Heaney LR (1991) A synopsis of climatic and vegetational change in Southeast Asia. Clim Change 19:53–61 References Hijmans R, Cruz M, Rojas E, Guarino L (2001) DIVA-GIS version 1.4: a geographic Alsos IG, Ehrich D, Eidesen PB, Solstad H, Westergaard KB, Schönswetter P, information system for the analysis of biodiversity data, manual. Interna‑ Tribsch A, Birkeland S, Elven R, Brochmann C (2015) Long-distance plant tional Potato Center, Lima, Peru dispersal to North Atlantic islands: colonization routes and founder effect. Hiramatsu M, Ii K, Okubo H, Huang KL, Huang CW (2001) Biogeography and AoB Plants 7:plv036 origin of Lilium longiflorum and L. formosanum (Liliaceae) endemic to the Barthlott W, Hostert A, Kier G, Küper W, Kreft H, Mutke J, Rafiqpoor MD, Som‑ Ryukyu Archipelago and Taiwan as determined by allozyme diversity. Am mer JH (2007) Geographic patterns of vascular plant diversity at conti‑ J Bot 88:1230–1239 nental to global scales (Geographische Muster der Gefäßpflanzenvielfalt Hsieh CF (2002) Composition, endemism and phytogeographical affinities of im kontinentalen und globalen Maßstab). Erdkunde 61:305–315 the Taiwan flora. Taiwania 47:298–310 Beck J, Böller M, Erhardt A, Schwanghart W (2014) Spatial bias in the GBIF Hsieh CF, Chao WC, Liao CC, Yang KC, Hsieh TH (1997) Floristic composition of database and its effect on modeling species’ geographic distributions. the evergreen broad-leaved forests of Taiwan. Nat Hist Res 4:1–16 Ecol Informatics 19:10–15 Hsieh CF, Chen ZS, Hsu YM, Yang KC, Hsieh TH (1998) Altitudinal zonation of ever‑ Bramwell D, Caujapé-Castells J (2011) The biology of island floras. Cambridge green broad-leaved forest on Mount Lopei, Taiwan. J Veg Sci 9:201–212 University Press, Cambridge Hsieh ML, Lai TH, Wu LC, Lu WC (2006) Eustatic sea-level change of 11–5 ka in Brooks TM, Mittermeier RA, Mittermeier CG, Fonseca GABD, Rylands AB, Western Taiwan, constrained by radiocarbon dates of core sediments. Terr Konstant WR, Flick P, Pilgram J, Oldfield S, Magin G, Hilton-Taylor G (2002) Atmos Ocean Sci 17:353–370 Habitat loss and extinction in the hotspots of biodiversity. Conserv Biol Hsu R, Wolf JHD (2009) Diversity and phytogeography of vascular epiphytes in 16:909–923 a tropical–subtropical transition island, Taiwan. Flora 204:612–627 Carlquist S (1967) The biota of long-distance dispersal. V. Plant dispersal to Huang JC (1996) Flora of Taiwan II. In: Hsieh CF (ed) Introduction to the flora of Pacific Islands. Bull Torrey Bot Club 94:129–162 Taiwan, 1: geography, geology, climate and soils, vol VI. National Taiwan Chao WC, Song GZM, Chao KJ, Liao CC, Fan SW, Wu SH, Hsieh TH, Sun IF, Kuo University, Taipei, pp 1–18 YL, Hsieh CF (2010) Lowland rainforests in southern Taiwan and Lanyu, at Huang SF (2011) Historical biogeography of the flora of Taiwan. J Natl Taiwan the northern border of Paleotropics and under the influence of monsoon Mus 64:33–63 wind. Plant Ecol 210:1–17 Huang S-F (2014) Hypothesizing origin, migration routes and distribution pat‑ Chen C-H (1957) Climatic classification and climatic regions. The Experimental terns of gymnosperms in Taiwan. Taiwania 59:139–163 Forest College of Forestry, National Taiwan University, Taipei Huang S-F, Lin T-P (2006) Migration of Trochodendron aralioides (Trochoden‑ Chen X-Y, He F-L (2009) Speciation and endemism under the model of island draceae) in Taiwan and its adjacent areas. Bot Stud 47:83–88 biogeography. Ecology 90:39–45 Huang SS, Hwang SY, Lin TP (2002) Spatial pattern of chloroplast DNA variation Chen YG, Liu TK (1996) Sea level changes in the last several thousand years, of Cyclobalanopsis glauca in Taiwan and East Asia. Mol Ecol 11:2349–2358 Penghu Islands, Taiwan Strait. Quat Int 45:254–262 Huang SF, Hwang SY, Wang JC, Lin TP (2004) Phylogeography of Trochoden- Chen ZS, Hsieh CF, Jiang FY, Hsieh TH, Sun IF (1997) Relationships of soil dron aralioides () in Taiwan and its adjacent areas. J properties to topography and vegetation in a subtropical rain forest in Biogeogr 31:1251–1259 southern Taiwan. Plant Ecol 132:229–241 Ingwersen P, Chavan V (2011) Indicators for the data usage index (DUI): an Chen C-H, Huang J-P, Tsai C-C, Chaw S-M (2009) Phylogeny of Calocedrus incentive for publishing primary biodiversity data through global infor‑ (Cupressaceae), an eastern Asian and western North American disjunct mation infrastructure. BMC Bioinformatics 12:S3 gymnosperm genus, inferred from nuclear ribosomal nrITS sequences. Kier G, Kreft H, Lee TM, Jetz W, Ibisch PL, Nowicki C, Mutke J, Barthlott W (2009) Bot Stud 50:425–433 A global assessment of endemism and species richness across island and Chiang TY, Schaal BA (2006) Phylogeography of plants in Taiwan and the mainland regions. PNAS 106:9322–9327 Ryukyu Archipelago. Taxon 55:31–41 Kokubugata G, Hirayama Y, Peng C-I, Yokota M, Möller M (2011) Phytogeo‑ Chiou C, Song G, Chien J, Wang J, Chen M, Liu H, Hsia Y, Chen T (2010) Altitudi‑ graphic aspects of Lysionotus pauciflorus sensu lato (Gesneriaceae) in nal distribution patterns of plant species in Taiwan are mainly determined the China, Japan and Taiwan regions: phylogenetic and morphological by the northeast monsoon rather than the heat retention mechanism of relationships and taxonomic consequences. Plant Syst Evol 292:177–188 Massenerhebung. Bot Stud 51:89–97 Kreft H, Jetz W, Mutke J, Kier G, Barthlott W (2008) Global diversity of island Chung K-F, Lin T-T, Tsai Y-S, Lin S-T, Peng C-I (2011) Isolation and characteriza‑ floras from a macroecological perspective. Ecol Lett 11:116–127 tion of microsatellite loci in Sassafras randaiense (Lauraceae). Am J Bot Krupnick GA, Kress WJ, Wagner WL (2009) Achieving target 2 of the global 98:e326–e329 strategy for plant conservation: building a preliminary assessment of Cowie RH, Holland BS (2006) Dispersal is fundamental to biogeography and vascular plant species using data from herbarium specimens. Biodivers the evolution of biodiversity on oceanic islands. J Biogeogr 33:193–198 Conserv 18:1459–1474 Denk T (2004) Revision of Fagus from the Cenozoic of Europe and southwest‑ Li C-F (2014) Diversity of mountain forest vegetation in Taiwan. Masarykova ern Asia and its phylogenetic implications. Doc Nat 150:S1–S72 Univerzita, Přírodovědecká fakulta, Brno Fang J, Song Y, Liu H, Piao S (2002) Vegetation-climate relationship and its Li J, Zhang D, Donoghue MJ (2003) Phylogeny and biogeography of Chamae- application in the division of vegetation zone in China. Acta Bot Sin cyparis (Cupressaceae) inferred from DNA sequences of the nuclear 44:1105–1122 ribosomal ITS region. Rhodora 105:106–117 Feng G, Mao L, Sandel B, Swenson NG, Svenning JC (2016) High plant end‑ Li CF, Chytrý M, Zelený D, Chen MY, Chen TY, Chiou CR, Hsia YJ, Liu HY, Yang emism in China is partially linked to reduced glacial-interglacial climate SZ, Yeh CL (2013) Classification of Taiwan forest vegetation. Appl Veg Sci change. J Biogeogr 43:145–154 16:698–719 Liao and Chen Bot Stud (2017) 58:15 Page 17 of 17

Liao CC, Chen YJ (2015) Fagaceae species of Taiwan relates to that of neigh‑ Saarenmaa H (2005) Sharing and accessing biodiversity data globally through boring regions and their distributions in East Asia. Taiwan J Biodivers GBIF 25th Annual ESRI International User Conference, San Diego, pp 17:205–221 25–29 Liao C-C, Liu M, Su M-H, Wang J-C (2013) Compression and overlap of unique Setoguchi H, Watanabe W, Maeda Y, Peng C-I (2008) Molecular phylogeny vegetation system of subtropical mountain resembling tropical and of the genus Pieris (Ericaceae) with special reference to phylogenetic temperate forests along elevation. J For Res 19:215–225 relationships of insular plants on the . Plant Syst Evol Liew PM, Hsieh ML (2000) Late Holocene (2 ka) sea level, river discharge 270:217–230 and climate interrelationship in the Taiwan region. J Asian Earth Sci Shen C-F (1997) The biogeography of Taiwan: 2. Some preliminary thoughts 18:499–505 and studies. Ann Rep Taiwan Mus 40:361–450 López-Pujol J, Zhang FM, Sun HQ, Ying TS, Ge S (2011) Centres of plant Su HJ (1984) Studies on the climate and vegetation types of the natural forests endemism in China: places for survival or for speciation? J Biogeogr in Taiwan (II) Altitudinal vegetation zones in relation to temperature 38:1267–1280 gradient. Q J Chin For 17:57–73 Mabry CM, Hamburg SP, Lin TC, Horng FW, King HB, Hsia YJ (1998) Typhoon Su HJ (1985) Studies on the climate and vegetation types of the natural forests disturbance and stand-level damage patterns at a subtropical forest in in Taiwan. (III).A scheme of geographical climatic regions. Q J Chin For Taiwan. Biotropica 30:238–250 18:33–44 Malcolm JR, Liu C, Neilson RP, Hansen L, Hannah L (2006) Global warming and Tang CQ, Chiou C-R, Lin C-T, Lin J-R, Hsieh C-F, Tang J-W, Su W-H, Hou X (2013) extinctions of endemic species from biodiversity hotspots. Conserv Biol Plant diversity patterns in subtropical evergreen broad-leaved forests of 20:538–548 Yunnan and Taiwan. Ecol Res 28:81–92 Matuszak S, Muellner-Riehl AN, Sun H, Favre A (2016) Dispersal routes between Teng LS (1990) Geotectonic evolution of late Cenozoic arc-continent collision biodiversity hotspots in Asia: the case of the mountain genus Tripteros- in Taiwan. Tectonophysics 183:57–76 permum (Gentianinae, Gentianaceae) and its close relatives. J Biogeogr Teng LS (2007) Quaternary tectonics of Taiwan. Spec Publ Cent Geol Surv 43:580–590 18:1–24 McCune B, Mefford MJ (2006) PC-ORD. Multivariate analysis of ecological data, Tiffney BH (1985) Perspectives on the origin of the floristic similarity between version 5. MjM software. Gleneden Beach eastern Asia and eastern . J Arnold Arbor 66:73–94 Nakamura K, Suwa R, Denda T, Yokota M (2009) Geohistorical and current Voris HK (2000) Maps of Pleistocene sea levels in Southeast Asia: shorelines, environmental influences on floristic differentiation in the Ryukyu Archi‑ river systems and time durations. J Biogeogr 27:1153–1167 pelago, Japan. J Biogeogr 36:919–928 Waldren S, Florence J, Chepstow-Lusty AJ (1995) Rare and endemic vascular Ni J (2003) Plant functional types and climate along a precipitation gradient in plants of the Pitcairn Islands, south-central Pacific ocean: a conservation temperate grasslands, north-east China and south-east Mongolia. J Arid appraisal. Biol Conserv 74:83–98 Environ 53:501–516 Wang W-T (1992a) On some distribution patterns and some migration routes Ohsawa M (1993) Latitudinal pattern of mountain vegetation zonation in found in the Eastern Asiatic Region (continue). Acta Phytotaxon Sin southern and eastern Asia. J Veg Sci 4:13–18 30:97–117 Ohsawa M (2006) Climate change impacts on vegetation in humid Asia moun‑ Wang W-T (1992b) On some distribution patterns and some migration routes tains. Global Environ Res 10:13–20 found in the Eastern Asiatic Region. Acta Phytotaxon Sin 30:1–24 Olson DM, Dinerstein E, Wikramanayake ED, Burgess ND, Powell GV, Under‑ Wei KY (2002) Environmental changes during the late Quaternary in Taiwan wood EC, D’amico JA, Itoua I, Strand HE, Morrison JC (2001) Terrestrial and adjacent seas: an overview of recent results of the past decade. West ecoregions of the World: a new map of life on Earth a new global map of Pac Earth Sci 2:149–160 terrestrial ecoregions provides an innovative tool for conserving biodiver‑ Wei X-X, Yang Z-Y, Li Y, Wang X-Q (2010) Molecular phylogeny and biogeog‑ sity. Bioscience 51:933–938 raphy of Pseudotsuga (Pinaceae): insights into the floristic relationship Ota H (1998) Geographical patterns of endemism and speciation in amphib‑ between Taiwan and its adjacent areas. Mol Phylogenet Evol 55:776–785 ians and reptiles of the Ryukyu Archipelago, Japan, with special reference Yesson C, Brewer PW, Sutton T, Caithness N, Pahwa JS, Burgess M, Gray WA, to their paleogeographical implications. Res Popul Ecol 40:189–204 White RJ, Jones AC, Bisby FA (2007) How global is the global biodiversity Padial JM, Miralles A, De la Riva I, Vences M (2010) Review: the integrative information facility? PLoS ONE 2:e1124 future of taxonomy. Front Zool 7:1–14 Zeng W-B (1993) The passage way of the flora migration on both sides of the Pianka ER (1966) Latitudinal gradients in species diversity: review of concepts. Taiwan Strait in Pleistocene Epoch. Acta Botanica Yunnanica 16:107–110 Am Nat 100:33–46 Zhang W, Lu Q, Liang J, Shen Z (2010) Altitudinal gradients of species richness Qian H, Ricklefs RE (2000) Large-scale processes and the Asian bias in species and range size of vascular plants in Taiwan: a test of Rapoport’s rule. diversity of temperate plants. Nature 407:180–182 Biodivers Sci 18:312–322 Qian H, Song J-S, Krestov P, Guo Q, Wu Z, Shen X, Guo X (2003) Large-scale Zhu H (1997) Ecological and biogeographical studies on the tropical rain forest phytogeographical patterns in East Asia in relation to latitudinal and of south Yunnan, SW China with a special reference to its relation with climatic gradients. J Biogeogr 30:129–141 rain forests of tropical Asia. J Biogeogr 24:647–662 Qian H, Fridley JD, Palmer MW (2007) The latitudinal gradient of species-area relationships for vascular plants of North America. Am Nat 170:690–701 Qiu Y-X, Fu C-X, Comes HP (2011) Plant molecular phylogeography in China and adjacent regions: tracing the genetic imprints of Quaternary climate and environmental change in the world’s most diverse temperate flora. Mol Phylogenet Evol 59:225–244